About this Blog
About The Author
Email Me

RealClearPolitics HorseRaceBlog

By Jay Cost

« Southern Democrats Feel Pressure from Obama Agenda | HorseRaceBlog Home Page | Obama's Tactical Mistake »

It's Not Just About Winning Elections...

Bill Greener has an article at Salon trumpeting the new enduring Democratic majority. We can now count him as yet another Republican to make this point. Greener has a reasonable conclusion, suggesting that the GOP needs to find a way to appeal to Hispanics without sacrificing its core principles. However, to make this argument he tells a tale about impending GOP doom that ultimately rests on some weak reasoning, most of which we have seen before. This point in particular struck me as especially flimsy:

In 1976, 90 percent of the votes cast in the presidential election came from non-Hispanic whites. In 2008, John McCain won this vote by a 56-43 margin. Had John McCain run in 1976 instead of 2008, not only would he have won, but he would have won the popular vote before a single non-white vote was cast.

What's the analytical purchase from re-running the 2008 vote with the 1976 demographics? How can you hold one constant and let the other vary? It seems to me that they are related - so if one changes, the other does, too. In particular, the Democrats have moved away from the South since Jimmy Carter ran in 1976; instead, they have looked to the West, the historical domain of the GOP. Isn't this at least in part due to the rise of Hispanics, especially in California? If so, how can you plug the new numbers into the old demographics? That will ultimately generate a bizarre result, like John McCain decisively winning an election that Jerry Ford couldn't. Er?

Also, this is just strange to me:

To make matters even worse, our weakness among minority voters is somewhat masked when it comes to elections that are not national. How can that be? Thanks to redistricting, and the legal imperative to give emphasis to "community of interests," these minority voters tend to be jammed into congressional districts where they are the overwhelming majority. That means the other districts tend to be more white in nature, and thus more friendly territory for Republicans. Then, at the level of the Senate, the reality is that Utah and Wyoming get the same number of senators as do California and New York.

What this means is that when it comes to an issue like immigration reform, the pressure on Republicans who actually have been elected to office is more often to favor a position that is unattractive to minority voters. If they were to take a different position, they might find themselves facing a primary challenger supported by the party's activist base. So, at the expense of any long-term perspective, the Republican Party is likely to be responsive to the sentiment of the people responsible for them serving at this very moment.

I'll grant that the racial distribution across congressional districts *might* "skew" legislative preferences on immigration (but see the footnote below for pushback on this point!) - although given the geographical concentration of Hispanics in just a handful of states, and then in discrete areas within many of those states, redistricting is not so much the problem as geographical-based single-member districts are. Still, this is actually a net electoral advantage for Republicans. Assume a 50-50 split among the parties, something akin to 2004. That year, George W. Bush won 255 congressional districts to Kerry's 180. Why the disparity? The Republican vote was distributed more evenly, while the Democratic vote was concentrated in urban and minority-majority districts. This is a distinct advantage that Republicans enjoy. To win the House, Democrats have had to win districts that Republican presidential candidates carry in 50-50 years. This is not inconsequential for public policy. As UCSD's Gary Jacobson noted in the Spring, 2009 issue of Political Science Quarterly, "Republicans hold a significant structural advantage in the competition for House seats," and:

This circumstance will have the effect of moderating the Democratic caucus, because Democrats representing such districts are, of political necessity, considerably more moderate than other Democrats. Similarly, more than half of the Democratic senators who replaced Republicans in 2006 and 2008 are from states in the South or the Mountain West, and they, too, will have to compile moderate records or risk defeat.

Hence Blue Dog resistance to health care reform. This also suggests that the Democrats might have the same problem in advancing immigration reform, given the large number of members who come from mostly white, conservative-tilting districts. You can stick either a Democrat or a Republican in districts like PA-12, MS-1 or CO-4. Those reps will be hard-pressed to vote yea. Perhaps this is why the issue has been tabled this year?

More broadly, assume that the distribution of GOP voters interferes with its ability to win Hispanics. Might not the Democrats have a similar problem? The chairmen of key House committees and other leaders - Frank, Rangel, Waxman, Conyers, Pelosi, etc - often come from districts that have little in common with swing districts. In fact, Bush's median share of the vote in 2004 in the districts of committee chairmen and leadership in the 111th Congress was just 36%. Can these Democrats be expected to have the individual incentives to craft policy designed to help the Democrats maintain a national majority?

This enduring majority argument also has some serious meta-level difficulties, too. Namely:

-You can't draw an inference about a trend (in this case, Hispanics, who looked to be shifting rightward in 2004) from a single data point.

-You can technically do this with two data points (in this case, young voters), but it can easily yield inferential errors. The Baby Boom generation was McGovern's biggest backer but ultimately voted for Reagan.

-You have to find a way to control for the fact that the economy was shrinking at a 6.1% annualized rate by Election Day; otherwise, movement that was induced by the economy (and other non-realigning factors) gets jumbled up with movement induced by realignment. I haven't seen any proponent of this hypothesis take a serious stab at controlling for the economy, as Sean Trende and I tried to do in our election reviews last winter.

-You can't lump African-Americans and Hispanics into a "non-white" group without ignoring the 3.67 million Hispanics who behaved contrary to the theory by...voting for John McCain! This lumping together of non-whites also ignores the volatility of Hispanic voters, who, as Greener does note, gave George W. Bush about 40% of the vote in 2004.

-You can't lump all Hispanics together as a single voting bloc - just as you can't do that for white voters, either! Some are more partial to the GOP than others. George W. Bush did relatively well with Hispanics in New Mexico (44%) and Nevada (39%), but poorly with Hispanics in New York (24%) and Illinois (23%). Additionally, the exit poll suggests that McCain actually did better with Hispanics in Colorado than George W. Bush.

The biggest problem of all is that the analysis is static. It fails to take into account whether and how the Democrats can hold their voting coalition together. It's not just about winning elections, it's about governance. The "realignment" of 1894/1986 was dependent not just on the shocks brought on by the harsh recession of those years, but also by the fact that the Republicans could govern to the satisfaction of the country, thus prohibiting Bryan and the Democrats from picking off enough voters from the Republican coalition in 1900. Today's arguments about a new, enduring Democratic majority are necessarily static because...the Democrats have not really governed yet! The age of Obama has just begun, seeing as we are only 12.5% of the way through his term.

An enduring majority requires victory across many elections, which in turn requires keeping enough of your voters happy. This is easier said than done - and Democrats have a challenge on their hands. The evidence of this is everywhere these days, and the highly astute Michael Barone connects a particular policy problem to the recent election:

Last Friday in the Beltway Confidential blog I called attention to the letter signed by Democratic Congressman Jared Polis and 20 other Democrats, 19 freshmen and one sophomore, opposing the $554 billion supertax on high earners included in the House Democrats' health care bill...

According to the Edison-Mitofsky exit poll, Obama carried voters with incomes under $50,000 and over $200,000. He lost among voters with incomes between $50,000 and $200,000. There's obviously a certain tension between high-income and low-income voters.

Incidentally, in Colorado, Congressman Polis's home state, the exit poll shows Obama getting a higher percentage (56%) among those earning more than $100,000 than among those earning less (53%). It shows Obama getting 53% from those between $100,000 and $150,000, so by interpolation those with incomes over $150,000 (who are the same percentage of the electorate as those in the $100,000-$150,000 bracket) cast 59% of their votes for Obama. Congressman Polis, who thanks to his success as an entrepreneur is among that high-income group, has evidently been paying attention. [Emphasis Mine]

There is a real tension here. Can the Democratic Party achieve its policy goals of improving the living standard of its lower-income voters without alienating its higher-income voters? I'm not so sure. Piling tax increase upon tax increase on the top earners could push them back to the GOP. Alternatively, avoiding tax increases yet running up larger and larger deficits might hurt the Democrats with middle income voters, whose budgets are such that they have to make the kinds of tough choices that (in this scenario) the Democratic-run government won't.

I'm not saying that there is no answer to this puzzle. It could be that the Democrats find a way to handle this tension, or calculate that they can side with one group over the other without losing their majority, or whatever. The point is that this is exactly the kind of balancing act majority parties must do, and that the challenge for the Democrats of actually governing should not be understated or ignored - which it inevitably is in these demographics-are-destiny arguments. As our President is now learning - railing against an unacceptable status quo on the campaign trail is substantially easier than altering it to the satisfaction of the electorate once in government.

The Democrats' majority will be enduring if and only if they can consistently satisfy enough of their voters amidst all of these challenges. Can they? We just don't know yet - and no amount of wishing by hopeful Democrats or fretting by hand-wringing Republicans can possibly change that fact.

* - Too much is made of the effects of gerrymandering on issue positions, and I am hesitant about going along with any conclusions like the one Greener makes connecting immigration to the gerrymander. Just the other day, I received the latest edition of the American Journal of Political Science, which contains an article written by Nolan McCarty, Keith Poole, and Howard Rosenthal. In it, they ask: "Does Gerrymandering Cause Polarization?" They answer: no! "[C]ongressional polarization is primarily a function of the differences in how Democrats and Republicans represent the same districts rather than a function of which districts each party represents or the distribution of constituency preferences." They "conduct simulations to gauge the level of polarization under various "neutral" districting procedures...[and] find that the actual levels of polarization are not much higher than those produced by the simulations." They conclude: "There are many reasons to do something about gerrymandering. But reducing polarization is not one of them."

-Jay Cost